mayko
International Hazard
Posts: 1218
Registered: 17-1-2013
Location: Carrboro, NC
Member Is Offline
Mood: anomalous (Euclid class)
|
|
ocean carbonate chemistry calculations
I am a geochemistry nerd with an interest in ocean acidification, but it has been an embarrassingly long time since I sat down and worked out a
carbonate system calculation.
Suppose that I know the temperature and atmospheric CO2 concentration at two points in time. I also know the ocean pH at one of these points in time.
I wish to know the change in pH.
Here's how I approached the problem. First I downloaded the SWCO2 package, which is a carbonate system calculator.
http://neon.otago.ac.nz/research/mfc/people/keith_hunter/sof...
Using the pH, pCO2, and temperature at point 1, I calculated a total dissolved inorganic carbon concentration (DIC, aka CT).
I then used Henry's law, with temperature-adjusted coefficients, to calculate the change in dissolved CO2. I added this change to the calculated DIC
for the first point. I then used the new DIC value, along with known temperature and pCO2, to calculate pH at the other point.
Am I doin' it right? My primary concern is when I simply add the change in dissolved CO2 to the DIC, which is the sum of the carbonate species in
equilibrium. It gets a ballpark correct answer, but is the calculation valid?
For those who are interested, I'm researching this article, perhaps to write a rebuttal. I am having a terrible time following the
calculations/reasoning; perhaps someone with more patience can decypher it?
http://wattsupwiththat.com/2013/07/08/ocean-acidi-what/
al-khemie is not a terrorist organization
"Chemicals, chemicals... I need chemicals!" - George Hayduke
"Wubbalubba dub-dub!" - Rick Sanchez
|
|
|
bfesser
|
Thread Moved
20-7-2013 at 09:28 |
mayko
International Hazard
Posts: 1218
Registered: 17-1-2013
Location: Carrboro, NC
Member Is Offline
Mood: anomalous (Euclid class)
|
|
While we're on the subject, other CO2 system calculators include
CO2SYS (Excell/MATLAB)
http://cdiac.ornl.gov/oceans/co2rprt.html
csys (MATLAB)
http://www.soest.hawaii.edu/oceanography/faculty/zeebe_files...
seacarb (R)
http://cran.at.r-project.org/web/packages/seacarb/index.html
al-khemie is not a terrorist organization
"Chemicals, chemicals... I need chemicals!" - George Hayduke
"Wubbalubba dub-dub!" - Rick Sanchez
|
|
|
mayko
International Hazard
Posts: 1218
Registered: 17-1-2013
Location: Carrboro, NC
Member Is Offline
Mood: anomalous (Euclid class)
|
|
I actually went ahead and contacted the author of the software; they had this to say:
| Quote: |
There is a simpler and more robust way. At point 1, use T, pH and pCO2 to calculate alkalinity. Then for point 2, use T, pCO2 and alkalinity to
calculate the pH. Alkalinity is independent of both T and changes in CO2 through gas exchange.
|
al-khemie is not a terrorist organization
"Chemicals, chemicals... I need chemicals!" - George Hayduke
"Wubbalubba dub-dub!" - Rick Sanchez
|
|
|
mayko
International Hazard
Posts: 1218
Registered: 17-1-2013
Location: Carrboro, NC
Member Is Offline
Mood: anomalous (Euclid class)
|
|
Full discussion posted here:
http://topologicoceans.wordpress.com/2013/08/14/whats-up-wit...
I found that SWCO2 was better for quick calculations, whereas the seacarb package was better for scripting. CO2SYS is the most popular program, but
I'm still a little confused as to how to use it.
al-khemie is not a terrorist organization
"Chemicals, chemicals... I need chemicals!" - George Hayduke
"Wubbalubba dub-dub!" - Rick Sanchez
|
|
|
bfesser
|
Thread Split
20-8-2013 at 04:53 |
bfesser
|
Thread Split
20-8-2013 at 05:42 |
AJKOER
Radically Dubious
Posts: 3026
Registered: 7-5-2011
Member Is Offline
Mood: No Mood
|
|
I am not an expert in this field, but my gut tells me this type of problem is better addressed by statistical modeling. You have dynamic interactive
systems subject to random errors with living organisms. Producing point estimated and statistically based confidence intervals on real data should be
important. I would try generalized least-squares recognizing that you may have missing variables and auto-correlation in the error terms. As such, the
data should be collected and modeled in time order.
I would recommend getting observations from a real bio-system that closely matches the subject ocean. Take repeated observations of pH as a function
of various variables (like temperature, measured in degrees Kelvin, and atmospheric CO2 concentration). Knowing the theoretic models will suggest the
appropriate transformation of the predictor variables (like employing the log transform for concentrations to be consistent with a dependent variable
of pH).
The problem I have with simulation models is that when a 'regime change' occurs (for example, a tipping point in temperature is reached and massive
amounts of CH4 are now, for the 1st time in recent history, being released) simulation models, not considering such new critical information, can be
highly unreliable and cannot provide any accurate risk measure. One can also perform a statistical test for a regime change itself. Knowing there is
such an event means that employing a long data series (for statistical analysis) is not appropriate as is a simulation model ignorant of this event.
[Edited on 8-12-2013 by AJKOER]
|
|
|
![[*]](images/xpblue/default_icon.gif){kind=icon}
